How do I translate binary to text?

[Togath]

So... I thought this would be a simple question, buuut google kinda failed for me. Is there any standard chart for translating binary to letters?
My curiosity got sparked while playing an old rpg(Breath of Fire 3) when this sequence popped up;

I mean, sure it miiight be random, but it made me curious if it actually said anything.

[Rockphed]

You would need to know what the encoding used is. I think ascii uses 7 bits per character, but you need at least 5 for a-z or 6 for a-Z. Alternatively, you could try decoding it as Morse code, but that also uses the time between pulses to tell when a character or word finishes, so it is ambiguous if you just have the dots and dashes.

[factotum]

You would need to know what the encoding used is. I think ascot uses 7 bits per character

You mean ASCII? That is usually a 7-bit encoding, yes, but the bit sequence in the original screenshot has 23 bits, which doesn't divide neatly into any sort of normal encoding scheme, unfortunately. My suspicion is that it's just a random sequence rather than having any sort of meaning.

[halfeye]

You mean ASCII? That is usually a 7-bit encoding, yes, but the bit sequence in the original screenshot has 23 bits, which doesn't divide neatly into any sort of normal encoding scheme, unfortunately. My suspicion is that it's just a random sequence rather than having any sort of meaning.

ASCII is a 7 bit encoding, but it would usually be encoded in 8 bits because memory is alway in powers of 2. Then there's Unicode in many versions, one of which is 8 bytes of 8 bits each which encodes pretty much every written human language including chinese.

Yeah, 23 is a prime number so that's got to be garbage.

[snowblizz]

You would need to know what the encoding used is.

Yeah. Without that information "binary" doesn't actually mean anything other than a number expressed using only 0 and 1.

Binary 10 stands for 2 in base10. 2 what? We don't know, could be number 2, could be 2nd character in a system, could be number 2 expressed in hexadecimal refers to a charactertable of some kind.

The OP's question is understandable, I'd be inclined to think it "meant" something too, but without further information than "it is binary" we cannot decode it.

[factotum]

ASCII is a 7 bit encoding, but it would usually be encoded in 8 bits because memory is alway in powers of 2.

That may be the case most of the time these days, but at the time ASCII was codified in 1963 there were all sorts of commonly used bit counts used as the basic size of a byte. In fact, the world's first integrated circuit microprocessor, the Intel 4004, had a 4-bit word, and I've heard of at least one machine (I forget which one) that used 14 bits. It wasn't really until the late 60s and early 70s that they started to standardise on 8 bits.

[Kaptin Keen]

As to how - well, you use a converter: https://www.rapidtables.com/convert/...-to-ascii.html

Whether it says anything, I couldn't say.

[jayem]

As to how - well, you use a converter: https://www.rapidtables.com/convert/...-to-ascii.html

Whether it says anything, I couldn't say.

If it's old there also telex, 5 bits iirc, various bastardisations of ASCII, and a=1 etc.
23 bits is still mysterious though (not impossible)

[Rockphed]

If it's old there also telex, 5 bits iirc, various bastardisations of ASCII, and a=1 etc.
23 bits is still mysterious though (not impossible)

The 23 bits is why I suggested morse code since different letters are different lengths.

[Peelee]

ascot

You mean ASCII?

Don't make a corny joke don't make a corny joke don't make a corny joke don't ma-

Spoiler: Corny joke!

Show

You know what they say, ascii a silly question, get a silly ansi!

[Rockphed]

Don't make a corny joke don't make a corny joke don't make a corny joke don't ma-

Spoiler: Corny joke!

Show

You know what they say, ascii a silly question, get a silly ansi!

*Slow clap.* Bravo sir. Bravo. If you didn't already have children I would say it was past time for you to reproduce. That level of punishment needs to be devoted to dad jokes.

[KillianHawkeye]

Given the character, who is a robot assistant, I assume it's probably R2D2-like beeps and boops rather than some kind of code.

[Aeson]

Yeah, 23 is a prime number so that's got to be garbage.

It's at least plausible that leading zeroes have been dropped, seeing as it's a displayed binary sequence and leading zeroes are not significant - assuming we're not messing around with a signed or fractional data type, anyways, though if you know the data type you could probably still understand it with leading zeroes dropped. Telling you that I have an eight-bit variable storing the value 00001011 isn't really any different than telling you that I have an eight-bit variable storing the value 1011, as long as it's understood that I'm dropping leading zeroes in the latter and giving you a binary number in both cases.

[Manga Shoggoth]

Don't make a corny joke don't make a corny joke don't make a corny joke don't ma-

Spoiler: Corny joke!

Show

You know what they say, ascii a silly question, get a silly ansi!

Since the leading bit on each line is 1, perhaps it is just paying her a complement?

[KillianHawkeye]

Since the leading bit on each line is 1, perhaps it is just paying her a complement?

Honey is the name of the robot, the one speaking, btw

[jayem]

Honey is the name of the robot, the one speaking, btw

It's a joke, 2's complement is a way of writing (negative) numbers in binary.

The problem with morse is (as already stated) you need to put the gaps in. And the way morse is designed makes that harder anyway.

There's base 64 (although that's a way of packing groups of 6 bits into groups of 80
I'm sure I have seen a morse like code where each letter end is obvious but the bit length of them is optimised (based around binary trees, unlike mores once you use a sequence for the letter that branch terminates).

[Manga Shoggoth]

It's a joke, 2's complement is a way of writing (negative) numbers in binary.

Yup - Binary is expressing ourselves in zeroes and puns.

(But I do appreciate KillianHawkeye's comment - I'm not familiar with the strip in question)

[factotum]

It's at least plausible that leading zeroes have been dropped, seeing as it's a displayed binary sequence and leading zeroes are not significant - assuming we're not messing around with a signed or fractional data type, anyways, though if you know the data type you could probably still understand it with leading zeroes dropped. Telling you that I have an eight-bit variable storing the value 00001011 isn't really any different than telling you that I have an eight-bit variable storing the value 1011, as long as it's understood that I'm dropping leading zeroes in the latter and giving you a binary number in both cases.

But that only works if you know the bit length of the pattern you're looking at. Is this 3x8 with leading zeroes dropped? 4x6? Also, a computer (or presumably the robot who this text is supposed to be from) never drops leading zeros from a bit stream, because the other end isn't going to know where the boundaries between the bytes are.

[Aeson]

But that only works if you know the bit length of the pattern you're looking at. Is this 3x8 with leading zeroes dropped? 4x6?

3x8 and 4x6 are the same length - 24 bits. Having the leading zero wouldn't help to distinguish either one of those from the other; a better example would be 4x7 and 3x8, since retention of the leading zeroes would allow you to distinguish between the 28-bit 4x7 stream and the 24-bit 3x8 stream.

Also, a computer (or presumably the robot who this text is supposed to be from) never drops leading zeros from a bit stream, because the other end isn't going to know where the boundaries between the bytes are.

If this were a real computer, that might be relevant, but it's a character in a video game whose "speech" is presumably written for human consumption by humans.

Also, the 'other end' can work out where the byte boundaries are, because unless you're a moron the only leading zeroes you'd drop are the first ones in the stream - you'd send 0001 0111 0101 1101 as 1011101011101, not as 11111011101. It'd be a bit of extra work as compared to just sending 0001011101011101, it'd have issues if for any reason the stream you needed to send lead with whatever was represented by the value 0 in whatever data type you're using, and for a real bit stream the reduction in signal length from dropping a few leading zeroes would probably be too insignificant to be worth the trouble of sorting out the boundaries at the far end, but it could be done.

Beyond that, while computers sending bit streams to other computers are very unlikely to be using a communications protocol that drops leading zeroes, computers displaying binary values for humans may drop the leading zeroes, because if it's being displayed for human perusal it'll often be formatted for human convenience before being put out; a 4x4 bit stream that goes 0001011101011101 could easily be displayed as 1 111 101 1101. Moreover, assuming computer-human audio communications using a two-tone system, there'd be greater incentive to drop leading zeroes than in a computer-computer bit stream, because the time savings for not sending the leading zeroes becomes much more significant - you probably can't send a person a continuous arbitrarily-long bit stream and expect them to interpret it correctly and the effective bit rate for a computer-human communications protocol in which the computer uses a two-tone audio signal to pass binary data to the human is necessarily going to be far lower than is typical for computer-computer communications protocols, so the time cost of retaining leading zeroes becomes much more significant, because you're breaking the bit stream up into smaller chunks so that the human can interpret them as it comes in (thus gaining more places in which it'd be reasonable to drop leading zeroes) and because signalling each bit takes significantly longer than would be the case for a computer-computer communication. Moreover, since humans do not normally communicate using two-tone binary audio signals, there's kind of a built-in assumption that the human is familiar with whatever protocol the robot is using to communicate and thus is reasonably likely to be able to interpret the message when leading zeroes are dropped.

[factotum]

Also, the 'other end' can work out where the byte boundaries are, because unless you're a moron the only leading zeroes you'd drop are the first ones in the stream - you'd send 0001 0111 0101 1101 as 1011101011101, not as 11111011101.

OK, just using an example from the ASCII character set--let's say I send a stream consisting of 110001. Is that the character 0110001 (e.g. number 1) or the character 110001 (e.g. letter b)? There's no way for you to tell if I've dropped the leading zero! I suppose you could count the total number of bits I send and figure out how many bits I dropped at the start from that, but that means you can't even start decoding the message until you've received it in its entirety. It's just computationally easier to handle if I send the entire bitstream including any leading zeroes--the minor saving in bandwidth usage isn't worth the massive additional complexity in decoding.

It's worth noting that in an actual transmission stream there will often be additional bits sent with every character so you know where they start and end, because if a couple of bits get corrupted or dropped due to transmission noise, you don't want to have to resend the entire thing.

[DataNinja]

Don't make a corny joke don't make a corny joke don't make a corny joke don't ma-

Spoiler: Corny joke!

Show

You know what they say, ascii a silly question, get a silly ansi!

I was expecting there to be a kernel of a joke in there. But all I saw was a husk of a pun.

[Altair_the_Vexed]

Of course, we're just assuming it's binary. It could be literally any base system, but coincidentally the number displayed has no other numerals in it.

[halfeye]

3x8 and 4x6 are the same length - 24 bits. Having the leading zero wouldn't help to distinguish either one of those from the other; a better example would be 4x7 and 3x8, since retention of the leading zeroes would allow you to distinguish between the 28-bit 4x7 stream and the 24-bit 3x8 stream.


If this were a real computer, that might be relevant, but it's a character in a video game whose "speech" is presumably written for human consumption by humans.

Also, the 'other end' can work out where the byte boundaries are, because unless you're a moron the only leading zeroes you'd drop are the first ones in the stream - you'd send 0001 0111 0101 1101 as 1011101011101, not as 11111011101. It'd be a bit of extra work as compared to just sending 0001011101011101, it'd have issues if for any reason the stream you needed to send lead with whatever was represented by the value 0 in whatever data type you're using, and for a real bit stream the reduction in signal length from dropping a few leading zeroes would probably be too insignificant to be worth the trouble of sorting out the boundaries at the far end, but it could be done.

Beyond that, while computers sending bit streams to other computers are very unlikely to be using a communications protocol that drops leading zeroes, computers displaying binary values for humans may drop the leading zeroes, because if it's being displayed for human perusal it'll often be formatted for human convenience before being put out; a 4x4 bit stream that goes 0001011101011101 could easily be displayed as 1 111 101 1101. Moreover, assuming computer-human audio communications using a two-tone system, there'd be greater incentive to drop leading zeroes than in a computer-computer bit stream, because the time savings for not sending the leading zeroes becomes much more significant - you probably can't send a person a continuous arbitrarily-long bit stream and expect them to interpret it correctly and the effective bit rate for a computer-human communications protocol in which the computer uses a two-tone audio signal to pass binary data to the human is necessarily going to be far lower than is typical for computer-computer communications protocols, so the time cost of retaining leading zeroes becomes much more significant, because you're breaking the bit stream up into smaller chunks so that the human can interpret them as it comes in (thus gaining more places in which it'd be reasonable to drop leading zeroes) and because signalling each bit takes significantly longer than would be the case for a computer-computer communication. Moreover, since humans do not normally communicate using two-tone binary audio signals, there's kind of a built-in assumption that the human is familiar with whatever protocol the robot is using to communicate and thus is reasonably likely to be able to interpret the message when leading zeroes are dropped.

If computers were sending to computers there might well be parity bits, I'm pretty sure I've heard of ASCII sometimes being 11 bits per byte.

Then again, there's also EBCDIC, which also codes binary to letters and numbers but is incompatible with ASCII (you can convert if you know which you have, but if you assume what you have is one but in fact it is the other, the result will be obvious garbage).

https://en.wikipedia.org/wiki/EBCDIC

[Aeson]

OK, just using an example from the ASCII character set--let's say I send a stream consisting of 110001. Is that the character 0110001 (e.g. number 1) or the character 110001 (e.g. letter b)? There's no way for you to tell if I've dropped the leading zero!

Computers - and, for that matter, humans - cannot distinguish between 0110001 (insignificant leading bit) and 0110001 (significant leading bit) without additional information, so your example is either wrong or incomplete. Since '1' is 0x31 (00110001) and 'b' is 0x62 (01100010) in the ASCII character set, I'm going to say it's wrong, because if you're only dropping leading zeroes then '1' is 110001 and 'b' is 1100010.

I suppose you could count the total number of bits I send and figure out how many bits I dropped at the start from that, but that means you can't even start decoding the message until you've received it in its entirety.

Which is a reason why it probably wouldn't be done in a real communications protocol, but we're probably not looking at an example of a real communications protocol in use.

It's just computationally easier to handle if I send the entire bitstream including any leading zeroes--the minor saving in bandwidth usage isn't worth the massive additional complexity in decoding.

Which is something I already mentioned in passing:

for a real bit stream the reduction in signal length from dropping a few leading zeroes would probably be too insignificant to be worth the trouble of sorting out the boundaries at the far end

If computers were sending to computers there might well be parity bits, I'm pretty sure I've heard of ASCII sometimes being 11 bits per byte.

11 bits per byte is more than just a parity bit, though I suppose it could be a parity bit for each nibble and a parity bit for the whole thing. With three bits added to every eight (or, less likely, four bits added to seven), it seems more likely that there's some form of error correction going on there, or at least a more complicated form of error detection than simple parity bits. Also, there are communications protocols where sending a single ASCII character would require significantly more bits than just 11 - for example, a very simple communications protocol with error correction is to send each bit in triplicate and correct transmission errors by majority vote at the receiving end.

Regardless, if we're looking at a bit stream produced by a communications protocol with error detection or error correction, then "it's 23 bits so it has to be garbage" is nonsense, but determining what it says requires knowing not only what data type we're looking at but also what communications protocol is in use.

[KillianHawkeye]

I've got to say that, if I were a game designer or script writer and I needed to have a small robot say something in "computer language", I'd just tap a bunch of ones and zeroes until I thought it looked good.

[georgie_leech]

I've got to say that, if I were a game designer or script writer and I needed to have a small robot say something in "computer language", I'd just tap a bunch of ones and zeroes until I thought it looked good.

It's not unheard of to put easter eggs in such places though. This one is far too short for, say, MADE YOU LOOK or something, but it's not the craziest idea either.

[factotum]

It's not unheard of to put easter eggs in such places though. This one is far too short for, say, MADE YOU LOOK or something, but it's not the craziest idea either.

Happened a few times with Bender in Futurama, as I recall. There's a horror-themed episode where he sees a binary code and isn't concerned about it, but when he sees it in a mirror, he freaks out--if you work out what the mirror code is, it's "666", aka the number of the beast.

[Peelee]

Happened a few times with Bender in Futurama, as I recall. There's a horror-themed episode where he sees a binary code and isn't concerned about it, but when he sees it in a mirror, he freaks out--if you work out what the mirror code is, it's "666", aka the number of the beast.

Using Futurama for an example of "not unheard of to put Easter eggs in" is almost cheating. There was a ridiculous level of attention to detail in that show.

[ereinion]

Googling some keywords related to the question in the OP, brought me to this GameFAQ page which claims to contain the game script to BoF3. In it, the binary is divided up into spaces:



That is "1001 0101 01110 10 00001 011.."

A cursory inspection of this doesn't reveal any further information to me, but maybe it'll help someone else here make more sense of it?

[jayem]

That is "1001 0101 01110 10 00001 011.."

Morse starts with PC or X depending on which way.
PC=a6d
which almost looks like something.